>In the early 20th century, both AC and DC electrification was used, with AC having the advantage of easier long distance transmission of power, and DC having the advantage that a train going downhill could use regenerative braking to pump current back into the wire from its traction motors (the same idea as dynamic braking on a diesel), so a downhill train could help power an uphill train.

Straight AC can be arranged for regerative braking too, without all of today's fancy electronics. It depends on the kind of [AC] traction motors in the locomotives and how the generating station and substation equipment is set up. The VGN and/or N&W used AC traction motors (can't recall at the moment whether they were single or polyphase), which will generate automatically when made to turn faster than "slip". When a train crested the summit, the locomotives started generating because the weight of the train speeded up the train slightly, driving the motors faster than slip. In a sense, the change over to regeneration was automatic. The generated power had to go somewhere, so it either went to another train or was dissipated with liquid rheostats at the generating and sub- stations. The need for dissipating the power also applies to DC lines like the Milw, which sent the regenerated power to another train or sold it back to the power company. If there is no load for the regenerated power, then there are no retarding effects. The PRR and NYNY&H didn't use regenerative braking.

>I'm not sure how that worked with motor-generator sets like the Milwaukee and other older DC operations used. If the train pumped DC back into the wire, it could turn the DC side of the M-G set as a motor, but if the AC side was not spinning at the right RPM to generate a 60 Hz current, you couldn't put a load on the AC side by connecting it to the grid.

The Milw locomotives were resistance machines; it was the GN (AC) locomotives which had M-G sets. With rotary conversion equipment, M-G sets and synchronous converters, located in either a locomotive or substation, there isn't much difficulty setting things up to send regenerated power back to the power company. Nor is there a problem with the frequency so generated because that is governed by the synchronous (or induction) motors of the M-G sets which are already synchronized with AC mains.

With rectifiers (mercury arc, ignitron tubes, and solid state), there has to be additional equipment to chop the DC into AC, smooth it, and make sure it's synchronized with the frequency of the power company's AC mains. Prior to solid state there was no cheap and efficient way to convert DC to AC at the voltages and current levels produced by regenerative braking without resorting to rotating conversion equipment, and the elimination of the expenses associated with rotating equipment was the big impetus for developing rectifiers.

For OPB: a little-known reason for the construction of the 750kv DC line between John Day and the LA area is that in the latter part of the 1960s some people in the USBR were all worked up that the USSR was ahead of the US in long-distance HV DC transmission systems. The reason "they" were ahead of "us" is because they had to generate power and move it several thousand km to the customers, with no customers along the line (like generate power in Salt Lake City and then send it to New Your City without selling power to St. Louis nor Chicago). The US has never had these distance problems, so we never bothered to develop the technology--at the time AC - HV DC and HV DC - AC conversion substations were very expensive, which was the main reason the US never got involved with the technology. In a real sense, the USBR people were doing a Chicken Little act, sort of like the missile and bomber gap affairs.